We first proposed that the IL-2 receptor was an interesting target for immunotherapy. Many workers have now validated the power of anti-CD25 mAb treatment in preclinical models. Five different anti-CD25 mAbs are immunosuppressive in clinical practice. Since rodent anti-human mAbs are poor cytocidal agents in man, we also proposed that the ideal therapeutic molecule for targeting the IL-2R would be a IL-2 toxin fusion protein. Our preclinical results support this hypothesis; however, we now aim to improve the potency of the IL-2 toxin-through protein engineering. It is critical that information be gained from these preclinical models before IL-2 toxin be employed as an immunosuppressive agent in man. In phase I-II clinical trials, the first IL-2 toxin has proven remarkably safe and potent in the elimination of IL-2R+ leukemic cells - even in patients bearing high titer, neutralizing anti-diphtheria toxoid antibodies. Unlike applications for cancer, the IL-2 toxin must compete for occupancy of the IL-2R with locally released IL-2 for successful treatment of GVHD. GVHD is resistant to the first generation IL-2 toxin. Our work outline follows: A. Protein engineering of IL-2-toxin. a. Analysis of DAB-IL-2 sequences that are required for maximal cytotoxic potency and optimal binding to the low, intermediate, and high affinity form of the IL-2 receptor.i. analysis of diphtheria toxin fragment B sequences between Thr387 and His485.i. deletion mutagenesis; ii. analysis of interleukin-2 sequences required for optimal binding to the high affinity form of the IL-2 receptor. i. site-directed mutagenesis b. Protein engineering of the region of the fusion junction between diphtheria toxin fragment B and human IL-2 sequences. i. insertion mutagenesis B. Determination of the potency of modified IL-2-toxin to cause 1. immunosuppression to conventional antigens 2. tolerance in murine GVHD models.